Finding lost animals

- A team Alpha project

Problem statment and target audiance

with our code and observations we tried different prediction models to see which one helps us predict the location of a lost pet . We also went one step ahead and tried making an user interface which takes in their house data and helps predict where the pet could have gone to. Our target audiance are pet owners whoes pets are inquisitive about wandering around the locality and end up getting lost on the way home .

Required Libraries

Loading Data

Understanding Animal Data

most of the animals were found close to the outcome address and were able to be returned

denisity of most animals are at shlter id 8 being dallas

Understanding shelter data

Observations from given data

From visualisations we understand

Mapping values to understand location effects

Only considering dallas data for understanding a small set

observations from maps

animals are found very rarley on a freeeway where are hugh ways and main roads have 1:5 chance of spotting an animal to be found from the surrounding houses

Predicting the location where the animals can be found

understanding linear regression model

because of the high error rate we shifted to regression tree models for more accurate predictions

Regression tree

Predicting for a single value input

Conclusion

We made a prediction model that helps us predict a certain location for a lost animal and map it using open street maps

Technologies and libraries

Future Scope

This project conclusions and final data points can also be re invested into making an application to notify a concerned pet owner the possible whereabouts of a lost pet and the near by shelters it can be found it .

Authors

Team Alpha -